As AI art generators become more publicly available, and chat bots become more advance, the question of AI gains more weight. If AI can generate great art, why have human painters? If AI can be lawyers, are human lawyers necessary? The question gains weight, anxiety builds, and those who oppose AI slip more and more into Luddism, universalizing their opposition to AI. Proponents of AI, save for Landians, doubt how far it can really go, thinking of AI, not as a super mind, but as a collection of 85 IQ interns that clear away menial task.
Behind the question there is a question. If AI, if algorithmic software, can replace humans, what are humans? Every question of AI has some idea of humanity implicit in the question. Is Monet the same type of being as Midjourney? For AI to be a threat, it has to be capable of replacing humans. For something to replace humans, that something has to be the same thing as humans, or capable of doing the same things humans do. I can only replace my truck, to illustrate, with something that does all the same things as my truck. This replacement must get me to and from work, have the same or better gas mileage, and be able to carry the same load. Sure, I could get a bicycle or a Prius, but this would not be strictly replacing my truck, it would be me being satisfied with something other than my truck. Maybe I decide I do not need a gasoline powered vehicle, or maybe I am okay not being able to move large loads anymore, and if I were to feel either of these ways, it would be me no longer wanting my truck or something analogous to it (like a Jeep). To suggest that AI could replace humans is, like replacing my truck, saying that AI is the same thing as, or can do the same things as, humans. What is being said when it is suggested that AI could replace Man?
Let us start with what AI is. AI is algorithmic software; it is designed by humans to complete defined tasks; its quality is measured by its computational power and analytical capabilities; it learns by reviewing preexisting data, and then using that data to accomplish its task(s). What is common to all of these qualities? We are here talking about purely intellectual qualities, although it would be more proper to say in some cases abilities. By “intellectual” qualities or abilities, I mean those qualities or abilities relating to discursive reasoning. Here we can already sense what kind of anthropology is assumed by the question of AI. If AI is intellectual, and if it can replace humans, then humans are intellectual beings. But before we make any conclusions, lest they be patchy, we must spend more time fleshing out AI.
Comparisons between AI and humans are almost exclusively between a computer and the brain. Sometimes the brain is likened to a computer, other time computers are likened to the brain. Points of comparison are 1) memory 2) computation and 3) pattern recognition. Note that, as said above, these three points of comparison are intellectual in nature. Relational aspects to the brain, not to mention the desiring faculty, is left out. Why? Maybe the capacity for relationships and the faculty of desire are found in the heart. Unless a robust theology of the heart, like that found in Saint Symeon the New Theologian, is in mind, it would be odd for STEM heavy folks to lean heavily into poetic device. Or maybe these capacities and faculties are seen as unimportant, asking if an AI can have a healthy relationship or if it wills to operate as it does might be asking the wrong question. For some reason, these questions are “irrelevant” and “about something” else, unrelated to the task at hand.
At this point, one can either dismiss the threat of AI, since it appears to be other than a relational and desiring being, and humans are such beings, or reaffirm its threat. Those that double down, giving recent advancements in AI as evidence of its threat, make, albeit unconsciously, a rather strong anthropological claim. The claim they make is an old one, going back to Augustine and Origen. In Analogical Identities: The Creation of the Christian Self, Father Nikolaos Loudovikos traces the influences of Augustine and Origen to Fredrich Nietzsche, arguing that “the will to power” is Christian in origin, while also exploring a rival conception of the self, championed by Maximus the Confessor, Gregory Palamas, and Symeon the New Theologian. Drawing from Father Nikolaos, we will paint the anthropology that opponents of AI assume, and then, by offering a rival conception of the self, nullify the threat of AI.
Loudovikos summarizes the Augustinian vision as follows: “For Augustine, a person is ‘a reasoning soul that uses an earthly body.’ This is the same definition that is to be found in Plotinus. Augustine’s spiritualism requires that human beings are fundamentally souls, that the soul is essentially the person. In this way, Augustine fits neatly into the philosophical tradition which begins with Plato and is completed in Neo-Platonism, although, as we shall see, he took human dualism much further, creating a new ontology of the ‘spiritual.’” From here, Loudovikos says, “In this way, Augustinian spiritualism achieved something without precedent in the history of human thought: regarding the soul, with consciousness at its centre, as the essence of the human person and as the image of God within him or her (with thought as its means of expression), Augustine was the first to move the locus of real life to thinking, connecting the will to it, as a vehicle of the ecstatic structure of thought.”1 The difference between Plato and Augustine is that, “whereas in Neo-Platonism the body is in real danger, in general, of ontological oblivion, here it remains, but is subject to the total authority of the soul. It remains under domination, with no power of its own, in thrall to an all-powerful ‘spiritual’ being, its soul.”2 It is the soul, and the soul as thinking, that is the real person. Yes, the body exists, and it was created good, but it is subject to the total authority of the soul. On the topic of creation, Loudovikos points out that “in the Augustinian interpretation of Genesis, the body was created after the soul, as Gilson mentions, and naturally, being subject to decay, does not bear the image of God.”3 Not only does the soul have vertical priority over the body (it has higher ontological priority), but it also has horizontal priority over body as well (it was created before the body). This is not to say Augustine thought the body was not real, or that it was evil, but that it is subject to the thinking soul and not the content of the self.
Augustinian spirituality, like Origenist spirituality, is predicated upon “the inward turn”, famously found in Book Ten, Chapters XIII-IX in Confessions. “The soul is the closed world of its meditation. And since nothing separates the thinking subject from the object it is thinking about, this unity is the fundamental certainty of Augustinian metaphysics and the only criterion of truth.”4 Turning inwards, finding the Kingdom of God promised to be in all of us, the soul is able to reflect upon its relation to the world, and to God, and this reflection is the process by which certainty is found. Note that the soul has no need for the body in its reflection, and not even the brain. Strangely, “in attempting to escape Manichaean sensualism and its concomitant materialism, Augustine denies almost any possibility of the senses influencing the soul. So, nobody ever learns anything. Everything already exists in the soul, not because the soul pre-existed, as Plato taught, but because it has an internal teacher which is forever instructing it. It has to do with a spiritual light, which is God Himself. Augustinian cogitation is simply the motion by which the soul gathers and arranges its as yet barely discernible knowledge in order to make it distinct.”5
Fleeing from his former heresy, Augustine prioritizes the soul to such an extent that it, and only its rational faculties, constitute the self. This view of the self is the same as a description of AI. Bodies are completely subject to the soul, bodies are not made in the image of God, and the soul’s essence is rational thought. To see AI as a threat requires adopting an Augustinian view of the self.
Before looking at a rival view of the self, one that is psychosomatic, I do want to raise the following question: “why would we want to adopt an Augustinian view of the self?” For those skeptical of metaphysics, it would seem odd to adopt a heavily metaphysical view of the self, locating the real self immaterially, and de-emphasizing the body. For Christians with a robust theology of the Eucharist, treating the body as the servant of the soul would also seem odd. To see AI as a threat, it would seem that the three following things are required 1) belief in the soul 2) the subjection of the body to the soul 3) a non-robust theology of the Eucharist. Only certain varieties of Protestants should be threatened by AI. Yet, the influence of Augustine in the West (and Origen in the East, who has a very similar view of the self) has been so far-reaching that many materialists and high church goers are unknowingly Augustinian. To counteract the influence of Augustine (or Origen) is a seemingly Symphisian task…at least with the pen. If a theology of the self was presented that nullified the threat of AI, it would open up the possibility of using AI as a tool, and having the latest technology at its disposal, this rival theology of the self would lay claim to future in the same way an electricity affirming theology would quickly outcompete an ant-electricity theology.
What is this rival theology of the self? Loudovikos begins with Maximus the Confessor and his conception of the will, “‘will’ (as θέλησις and not βούλησις) is, in the first place, a universal, non-metaphysical, personal and profoundly direct desire to open up to the complete Being. Indeed, while the tradition before him abandoned the will, (as βούλησις), as prey to the intellect or the passions (or to a conflict between the two, he existentializes the will (as θέλησις), as ‘a simple, reasonable and vital appetite,’ in contrast to the (for example Aristotelian) προαίρεσις, deliberate choice, which is not, of course, a passive attraction to an object, but a mental, metaphysical attraction, intellectual and abstract, as ‘the companion of appetite, deliberate choice and judgement.’”6 Unlike the thought that came before him (which includes Augustine and Origen), Maximus gives us something that looks a lot like the unconscious. Will is not the intellect freely choosing an object, but our inmost desire to open up to God. It is "through its will, [that] the being seeks the full entity of its nature, the complete ‘according to nature’ realization of all its essential component parts."7 Nature and will are unopposed. We will to be whole, and our whole self is how God created us to be. Opposition between God's providence and our will, found in the Reformed tradition, is absent in Maximus because of his break with an intellectual definition of will. Will, to repeat, is not my intellect choosing some action after deliberation, but my inmost desire to be myself. To clarify this, and to see its psychosomatic dimension, let us quote Loudovikos again,
“The uncreated, volitional, affective proposition which gives substance to beings, as the reason of Being, is a divine eschatological proposition to them, not a teleological destination for them, such as the Platonic ideas are. This is why beings are linked to their principle ‘by appetite,’ as a desire for a process of becoming which completes their truth eschatologically by opening them up fully towards it. The introduction of the will as a θέλημα, a burning desire beyond reason, encompassing not only reason but also sentiments, desires, the bodily dimension, and also perhaps what a modern psychologist would call the unconscious, into ontology makes it eschatological, then; in other words makes it the ontology of the freedom of nature: nature ‘becomes,’ through a free desire, and in dialogue with the divine θέλησις/desire, and it not merely ‘is.’...This understanding of the will, as natural will, which expresses the internal life of the whole nature, as a burning desire for freedom and incorruption..."8
I, by my natural will, which is not the intellectual will found before Maximus, desire to be whole, and being whole means that all parts of me, my reason, sentiments, desires, and body is in open dialogue with God. All parts of me are whole when all parts of me are as God wants them to be, when they serve their vocational function. Note the psychosomatic nature of this idea. There is, for Maximus, a vocation for my body, and because of this the body is as much as my real self as my reason, and the same goes for sentiments and desires. Being a parent, lumberjack, cook, dancer, and musician, these are all bodily vocations, and, for those called to them, they help constitute the self. When we become whole, this is what the Orthodox tradition calls “theosis”, or deification.
Unfortunately, many mistake theosis to be the eclipsing of our nature, and a type of spiritual will to power by which, through prayer and fasting, I can overcome my nature and achieve the rank of a god. However, this is far from the patristic witness,
“Deification does not mean an outlet and abandonment of the natural, but its eschatological completion. As opposed to Pyrrhus, who identified nature with necessity, as did the average intellectual of his day (and ours), Maximos counter-proposes, most paradoxically, that not only does the divine nature have nothing natural which is subject to necessity, but that this is true also of created, rational nature! Why is this? Precisely because created rational nature has will as the character of its Being, i.e. it constitutes its Being, as the eschatological process of growing into healing the rifts that the Fall has created, breaking nature in fragments of hostile gnomic wills, opposing each other, through communion with the uncreated, which frees from this fragmentation: hence ‘the natural characteristics of the intellectual beings are not bound with necessity!”9
Without diving into what “gnomic wills” are, which would take us too far astray, we can simply say that theosis is the healing of the rifts created by the Fall. Our nature is not transcended but restored.
Contrasted with the Augustinian (and Origenist) view of the self, Maximus’ view affirms our sentiments, desires, and body to be as real as our reason, and the totality of these make up the self. No part is subject to any other, because all are called by God. It is not only my reason, by its very nature, that craves communion with God, but behind every sentiment of mine is craving for communion. My very body, and this I find fascinating, as body, not as poetic device, craves, in its inmost being, to be whole as it was prior to the fall. Theosis is the psychosomatic healing of our Fallen nature.
AI, not matter how advanced, cannot hold a candle to Man. Assuming its reason is equal to ours (which it is not, since a computer does not naturally will communion with God), AI lacks all the other constituent parts to the Maximian self. A follower of The Confesor has no problem with AI, and can see it as a helpful tool.
Behind the question of AI is the question of the self, who Man really is. Only if we assume a very particular view of the self, is AI a threat. We are not, despite the popular claim, forced to choose between Luddism and being replaced. AI cannot replace us any more than a hammer can. There may be some jobs automated way, yes, but this will be similar to any other period of human history where jobs were automated. Most shoes are made by machines, not cobblers. Those that still are, have become luxury items. It is true that the economy will likely change some, but it is not true that humans will be replaced. There will still be teachers, doctors, lawyers, and other human professions. AI can be a helpful tool, so long as we have a sound theology of the self.
Loudovikos, Nikolaos. Analogical Identities: The Creation of the Christian Self. Brepols Publishers. Turnhout, Belgium. 2019. 21
Loudovikos, Nikolaos. Analogical Identities: The Creation of the Christian Self. Brepols Publishers. Turnhout, Belgium. 2019. 21 21
Loudovikos. 22
Ibid. 25
Ibid. 25
Ibid. 69
Ibid. 70
Ibid. 70
Ibid. 72
'AI can be a helpful tool, so long as we have a sound theology of the self.'
Let me try to put this quote in perspective, by replacing 'AI' with 'sex robots' and 'theology of the self' with 'human relationships':
'Sex robots can be a helpful tool, so long as people favour healthy human relationships.'
Why? Because robots cannot really fulfil us; the self is so much more than a computer etc.
The real question is: can techno-optimists *replace* healthy human relationships with robot surrogates? Ever heard of porn hub?
It was very interesting to read about Augustine vs Maximus and I agree with the definition of self as psychosomatic. However, to jump from this to 'AI is therefore not a threat' seems like dismissing the essence of its disruptive effects.
As an architect and digital artist, my main concern is the role of AI text-to-image art generators like Midjourney or Stable Diffusion.
A good summary of their pernicious consequences is given by Steven Zapata in a manifesto that can be found on YouTube under the title 'The End of Art: An Argument Against Image AIs'.
We are not concerned that AI might develop better thinking minds than humans, we are concerned that:
1) human decision-makers might advance false definitions of human nature, instantiate them into laws and customs and force societies to comply to them. Patrick Deneen's 'Why Liberalism Failed' provides ample evidence of this happening over the past 300 years
2) human innovation developing chaotically, indifferent to *any* definition of human nature, community, telos
These two phenomena have been happening since the industrial revolution (are we uncle-Ted-pilled?) and it can be argued that ethicists and scientistic-minded people have been providing post-hoc rationalisations of unleashed technology; they either sold it as 'progress', or as an equalising force, or reduced all higher order values (Scheler's pyramid) to utility and pleasure.
With AI we have all these already existing drives, ramped up to 11. AI art generators were created by learning from copyrighted material (Zapata's video covers this), and are now used as a replacement for human artists. I don't care whether they could do a better job, this is immaterial and of a metaphysical nature; the concrete reality is CEOs of large animation or architecture companies (like Zaha Hadid) starting to generate digital art without paying artists to do it, but rather learning how to carefully type words in the text box of Midjourney, resulting in images that later inform the company's designs.
It is strange that in the 20th century, the justification of automatising everything was the elimination of hard labour and unpleasant chores, so that men could spend time doing pleasant activities like *art*. Now it seems even that is taken from him.
Last but not least, AI art has a cheapening effect and a further destruction of the aura of authenticity (Walter Benjamin wrote a good essay on the influence of mass production in eroding the aura of art); with NFTs and AI generative models, this process reaches unimaginable depths. We already know that facebook, google and amazon collect our preferences and sell them to 3rd parties that can then offer targeted ads, or political parties can recruit, watch us or shadowban us. When this data is connected with AI art generators, you will start seeing content created instantly targeted at individuals; imagine an AI-generated comic book that develops an idea you had, but never managed to find the time to materialise it; or a music video about the death of your pet, or the nasty words your mother said to you. It is happening, Stable diffusion are already working on creating an AI music creator.
A technology does not have to develop a mind of its own in a true metaphysical sense to cease being a tool; disruptive technologies *are* becoming strange golems and fetishes that force us to adapt or perish, kinda like in an arms race.
"There is really no reason why a human being should do more than eat, drink, sleep, breathe, and procreate; everything else could be done for him by machinery. Therefore the logical end of mechanical progress is to reduce the human being to something resembling a brain in a bottle. That is the goal towards which we are already moving, though, of course, we have no intention of getting there; just as a man who drinks a bottle of whisky a day does not actually intend to get cirrhosis of the liver" - Orwell, The Road to Wigan Pier